Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print January 14, 2008, 10.1148/radiol.2463070293
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow AVI Movies
Right arrow All Versions of this Article:
2463070293v1
246/3/871    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Taschner, C. A.
Right arrow Articles by Leclerc, X.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Taschner, C. A.
Right arrow Articles by Leclerc, X.
(Radiology 2008;246:871-879.)
© RSNA, 2008

Neuroradiology

Intracranial Arteriovenous Malformation: Time-resolved Contrast-enhanced MR Angiography with Combination of Parallel Imaging, Keyhole Acquisition, and k-Space Sampling Techniques at 1.5 T1

Christian A. Taschner, MD, Jürgen Gieseke, PhD, Vianney Le Thuc, PhD, Henda Rachdi, MD, Nicolas Reyns, MD, Jean-Yves Gauvrit, MD, and Xavier Leclerc, MD

1 From the Departments of Neuroradiology (C.A.T., V.L.T., H.R., J.Y.G., X.L.) and Neurosurgery (N.R.), Hôpital Roger Salengro, University Hospital Lille, Rue Emile Laine, F-59037 Lille Cédex, France; and Philips Medical Systems, MR Clinical Science, Best, the Netherlands (J.G.). Received February 12, 2007; revision requested April 13; revision received June 12; accepted July 18; final version accepted August 30. C.A.T. supported by a research grant from the Swiss National Science Foundation that was funded by the L. + Th. Laroche Stiftung, Basel, Switzerland. Address correspondence to C.A.T. (e-mail: c.taschner{at}web.de).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Purpose: To prospectively compare the agreement between digital subtraction angiography (DSA) and time-resolved magnetic resonance (MR) angiography with sensitivity encoding (SENSE) in combination with keyhole acquisition and contrast material–enhanced robust-timing angiography (CENTRA) k-space sampling techniques for the characterization of intracranial arteriovenous malformations (AVMs).

Materials and Methods: The institutional review board approved the study; informed consent was obtained from all patients (or their parents). Twenty-eight patients (15 male, 13 female; mean age, 38.6 years; age range, 16–61 years) with 29 previously diagnosed, untreated intracranial AVMs who were referred for stereotactic gamma knife radiosurgery were evaluated. Preinterventional imaging included intraarterial DSA and time-resolved MR angiography. The time-resolved MR angiography sequence included SENSE with a 1.5-T imager and was optimized by applying keyhole acquisition and CENTRA techniques. Time-resolved MR angiograms were reviewed by two independent raters and compared with DSA images with regard to arterial feeders, nidus size, and venous drainage. {kappa} Statistics were applied to determine interobserver and intermodality agreement.

Results: MR angiography enabled time-resolved (1.7 seconds per volume) visualization of cerebral vessels from axis to vertex at high spatial resolution (true voxel size, 1 x 1 x 2 mm). All 25 nidi detected at intraarterial DSA were visualized at time-resolved MR angiography. Intermodality agreement was excellent for arterial feeders ({kappa} = 0.91; 95% confidence interval [CI]: 0.786, 1.000) and venous drainage ({kappa} = 0.94; 95% CI: 0.814, 1.000) and was good for nidus size ({kappa} = 0.76; 95% CI: 0.562, 0.950).

Conclusion: The agreement (good to excellent) between time-resolved MR angiographic and DSA findings suggests that time-resolved MR angiography is a reliable tool for the characterization of intracranial AVMs with respect to arterial feeders, nidus size, and venous drainage.

Supplemental material: http://radiology.rsnajnls.org/cgi/content/full/2463070293/DC1

© RSNA, 2008


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Intraarterial digital subtraction angiography (DSA) remains the reference standard for the assessment of intracranial arteriovenous malformations (AVMs). Because of its inherent high spatial and temporal resolution, DSA allows accurate analysis of nidus angioarchitecture, with separation between feeding arteries and draining veins. However, this examination is invasive and requires a radiation dose and iodinated contrast material injection.

Alternative techniques have been proposed for the visualization of the intracranial vasculature to avoid the use of DSA. Several groups (1–3) initially suggested time-of-flight magnetic resonance (MR) angiography for the visualization of AVMs. However, the clinical use of this technique remains limited because of its long imaging time, which prevents accurate differentiation between arterial and venous phases at MR angiography. Moreover, saturation effects inherent to the technique may be responsible for decreased signal intensity in capillaries and veins. Substantial technologic advances have been made with time-resolved, contrast material–enhanced MR angiography with complex subtraction. This method combines the T1 shortening effect of a gadolinium-based contrast agent, subsecond imaging, and the digital subtraction technique, which allows high-temporal-resolution MR angiography. Results of preliminary studies (4–7) have shown the potential interest in this technique for the assessment of cerebrovascular pathologic conditions. More recently, the use of parallel imaging techniques, such as sensitivity encoding (SENSE), has considerably accelerated image acquisition at time-resolved MR angiography (8–13). Alternatively, temporal and spatial resolution can be improved by applying intelligent k-space sampling techniques (14).

Farb et al (15) showed that the use of k-space sampling with elliptic centric-ordered phase encoding could improve the image contrast and vessel conspicuity resulting from fast MR angiographic sequences. A combination of keyhole acquisition with segmented central k-space ordering (ie, contrast-enhanced robust-timing angiography [CENTRA]) techniques has been reported to improve temporal resolution at contrast-enhanced MR angiography at a constant spatial resolution (16,17).

The aim of our study was to prospectively compare the agreement between DSA and time-resolved MR angiography with SENSE in combination with keyhole acquisition and CENTRA k-space sampling techniques for the characterization of intracranial AVMs.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
An employee (J.G.) of Philips Medical Systems (Best, the Netherlands) assisted in MR imaging sequence development and imaging protocol design. The other authors, however, had full control of the data and information submitted for publication.

Patients
Our study was approved by our institutional review board, and informed consent was obtained from all patients or from the parents of any underage patient. Our prospective study included 28 patients (age range, 16–61 years; mean age, 38.6 years ± 12.4 [standard deviation]) with 29 previously diagnosed cerebral AVMs who were referred for gamma knife radiosurgery between October 2005 and June 2006. Our study was restricted to patients with untreated AVMs. Thirteen patients were women (age range, 22–58 years; mean age, 38.4 years ± 10.9), and 15 patients were male (age range, 16–61 years; mean age, 38.8 years ± 14.1). Patients were referred for DSA and MR imaging, including time-resolved MR angiography, prior to gamma knife treatment. The interval between DSA and MR angiography ranged from 6 to 24 hours. Time-resolved MR angiography was performed before radiosurgery in all patients. Time-resolved MR angiography served as a baseline examination for patient follow-up after radiosurgery.

According to results of the initial diagnostic DSA examination performed at peripheral hospitals, 25 AVMs were located supratentorially (two frontal, eight parietal, three frontoparietal, six temporal, four occipital, two parieto-occipital), and four AVMs were located infratentorially. The presenting symptoms of the underlying AVMs were intracranial hemorrhage in 12 patients, headaches in eight, and seizures in six. In two patients, the AVMs were discovered incidentally (Table 1).


View this table:
[in this window]
[in a new window]

  		Table 1. Summary of Patients and Cerebral AVMs

 
DSA Technique
The time between the diagnosis of AVM and the scheduled treatment by using gamma knife radiosurgery ranged between 6 and 20 months. DSA (Allura; Philips Medical Systems) was performed in stereotactic conditions on the day of radiosurgery by one of two trained neuroradiologists (H.R., with 7 years of experience performing DSA, and X.L., with 20 years of experience with DSA). DSA included a selective injection in the common carotid or vertebral arteries with a 4-F catheter by using a femoral artery approach. Frontal and lateral views were obtained, and a filming rate of three images per second was used. In our experience, this filming rate is sufficient to reliably define the margins of the nidus of the AVM for dosimetry planning of stereotactic gamma knife radiosurgery. Yet for diagnostic purposes and to identify aneurysms of the feeding arteries or within the nidus, higher filming rates might be necessary. Images were obtained with a 512 x 512 matrix and a field of view of 17 cm. For each projection, an 8–16-mL bolus of an undiluted iodinated contrast material with an iodine concentration of 250 mg/mL (iobitridol, Xenetix; Guerbet, Aulnay-sous-Bois, France) was injected at a rate between 3 and 6 mL/sec by using a power injector.

MR Imaging
Time-resolved contrast-enhanced MR angiography was performed without a stereotactic frame by using a 1.5-T MR system (Achieva; Philips Medical Systems) with a commercially available eight-channel head coil. The MR unit was equipped with a gradient system that allowed a maximal achievable gradient amplitude of 30 mT/m, a rise time of 0.2 msec, and a slew rate of 150 T/m/sec.

Patients were positioned with a 20-gauge intravenous catheter inserted into the antecubital vein. Intravenous injection of 0.2 mL of gadoterate meglumine (Dotarem; Guerbet) per kilogram of body weight at a flow rate of 6 mL/sec was followed by a 10-mL saline flush by using an automated power injector (Spectris; Medrad, Indianola, Pa). Time-resolved MR angiography was initiated after injection of the contrast agent. The acquisition parameters of the time-resolved MR angiographic three-dimensional (3D) T1-weighted gradient-echo sequence were as follows: repetition time msec/echo time msec, 3.64–3.85 (mean, 3.70 ± 0.615)/1.3; flip angle, 15°; image matrix, 256 x 256; and a field of view of 256 mm covering the entire head.

One hundred sixty-five thin sagittal partitions of 2 mm with a 1-mm overlap between sections were acquired with a SENSE factor of two in the section-selection direction, a SENSE factor of four in the phase-encoding direction, and a keyhole diameter of 16%, which yielded a near-isotropic true voxel size of 1 x 1 x 2 mm (2 mm3) without zero filling. In total, 13 dynamic volumes were acquired with an average keyhole imaging duration of 1.7 seconds per volume, followed by dynamic reference imaging with a duration of 10.9 seconds. Total acquisition time for the time-resolved MR angiographic sequence was 33 seconds. Time-resolved MR angiography yielded a total acceleration factor of 66.6 (6.25 [16% keyhole] · 8 [SENSE])/0.75 [half scan factor]) as compared with standard contrast-enhanced MR angiography without such techniques. Image processing included mask subtraction to suppress the background signal of the stationary tissue. For this purpose, we used one of three dynamic volumes acquired prior to administration of the contrast agent with the same time-resolved MR angiographic sequence (imaging duration = 16 seconds). The total imaging duration of time-resolved MR angiography was 50 seconds. Images were displayed with a reconstructed isotropic image matrix of 1 x 1 x 1 mm. In addition, routine MR imaging included pre- and postcontrast T1-weighted spin-echo and T2-weighted fast spin-echo sequences. Time-resolved contrast-enhanced MR angiography with parallel imaging in combination with keyhole acquisition and CENTRA k-space sampling techniques was successful in 28 of 28 patients.

Image Analysis
Two independent raters (H.R. and X.L., with 7 and 20 years of experience in neuroimaging, respectively) qualitatively evaluated all DSA images on hard copies. Studies that led to disagreement were reviewed to reach consensus. Two other raters (C.A.T. and J.Y.G., with 9 and 12 years of experience in neuroradiologic MR imaging, respectively), who were blinded to clinical and DSA results, independently evaluated the time-resolved MR angiographic data at a computer workstation (Easy Vision; Philips Medical Systems) and applied maximum intensity projection display algorithms. The 3D data were displayed with all regions visible. The software enabled the enlargement of regions of special interest in any given spatial orientation. Raters evaluating the time-resolved MR angiographic data were informed that spontaneous regression of the previously diagnosed AVM might have occurred.

Time-resolved MR angiographic and conventional DSA data were assessed for nidus size, arterial feeders, and venous drainage of the underlying AVM. Arterial feeders were defined as branches deriving from the anterior cerebral artery, middle cerebral artery, posterior cerebral artery, or a cerebellar artery. AVMs were stratified with respect to nidus size as small (<1 cm), medium (1–3 cm), or large (>3 cm). The drainage of the AVMs was recorded as being superficial, deep, or deep and superficial, where superficial was defined as any drainage into the cortical venous system and deep was defined as drainage occurring through one of the veins of Galen, the straight sinus, or the basal veins. For studies in which the two raters disagreed, consensus reading of the time-resolved MR angiographic data was performed and noted separately.

Statistical Analysis
The levels of interobserver agreement (between rater 1 and rater 2 for time-resolved MR angiograms) and intermodality agreement (between consensus readings of time-resolved MR angiograms and DSA images) with respect to arterial feeders, nidus size, and venous drainage were determined by calculating the {kappa} coefficient ({kappa} < 0.20 indicated poor agreement; {kappa} = 0.21–0.40, fair agreement; {kappa} = 0.41–0.60, moderate agreement; {kappa} = 0.61–0.80, good agreement; {kappa} = 0.81–0.90, very good agreement; and {kappa} > 0.90, excellent agreement). In addition, the exact number and percentage of times that results from the two raters and the two modalities were in exact agreement were provided, including exact and 95% confidence intervals (CIs). A statistical package (MedCalc; MediSoftware, Mariakerke, Belgium) was used to perform all calculations. While data were available for 29 AVMs in 28 patients, data from only one AVM (chosen at random) were used for the patient with two AVMs.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
We obtained a near-isotropic whole-brain 3D data set with a spatial resolution of 1 x 1 x 2 mm (2 mm3) without zero filling at a frame update time of 1.7 seconds per volume, which allowed multiple reconstructions of time-resolved MR angiographic data in any spatial orientation in all patients (Figs 1, 2).


Figure 1
View larger version (87K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Maximum intensity projection of time-resolved MR angiographic data (3.7/1.3, 15° flip angle) reconstructed in sagittal plane in 16-year-old patient with AVM in left parietal lobe shows passage of contrast agent from early arterial (top left) to late venous phase (bottom right). Projection displays a plexiform nidus (long arrow) in left parietal lobe, supplied by branches deriving from anterior cerebral artery (short arrow). Venous drainage occurs via a dilated cortical vein (arrowhead) into superior sagittal sinus. (See Movies 1 and 2, http://radiology.rsnajnls.org/cgi/content/full/2463070293/DC1).

 

Figure 2
View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2: Oblique reconstructions of 3D volume acquired during parenchymal phase at time-resolved MR angiography in 31-year-old woman with AVM in left parietal lobe. Near-isotropic volume data allow image reconstruction in any given projection. Volume acquired during parenchymal phase at MR angiography (3.7/1.3, 15° flip angle) and reconstructed in five oblique projections between 0° and 90° (I–V) displays nidus of AVM (long arrow), arterial feeders (short arrow), and draining veins (arrowhead). (See Movies 3 and 4, http://radiology.rsnajnls.org/cgi/content/full/2463070293/DC1).

 
Of 29 previously diagnosed AVMs, 25 were present at DSA on the day of radiosurgery. Four small AVMs could not be visualized at DSA. This might be related to spontaneous regression of the underlying AVM within the time between the initial diagnosis of the AVM and the scheduled treatment by using radiosurgery, which ranged between 6 and 20 months. All 25 AVMs detected at DSA were visualized at time-resolved MR angiography, and no additional AVMs were found at time-resolved MR angiography (Table 2). We derived a single observation for statistical analysis for the patient with two intracranial AVMs.


View this table:
[in this window]
[in a new window]

  		Table 2. Summary of Intraarterial DSA and Time-resolved MR Angiographic Findings in 28 AVMs

 
Time-resolved MR angiography (consensus reading) and DSA were in agreement for the major arterial branches supplying the nidus of the AVM in 25 (89%) of 28 studies (exact CI: 0.718, 0.977). At DSA, four AVMs were supplied by the anterior cerebral artery; 11, by the middle cerebral artery; four, by the posterior cerebral artery; two, by the anterior cerebral artery and the middle cerebral artery; one, by the middle cerebral artery and the posterior cerebral artery; and two, by cerebellar arteries. Intermodality agreement (consensus reading of time-resolved MR angiograms vs DSA images) was excellent ({kappa} = 0.91; 95% CI: 0.786, 1). The analysis of the arterial supply of AVMs showed agreement in 26 (93%) of 28 studies (exact CI: 0.765, 0.991) between the raters reviewing time-resolved MR angiograms, which resulted in excellent interobserver agreement ({kappa} = 0.91; 95% CI: 0.790, 1.000).

The evaluation of nidus size showed a greater variability between techniques. Time-resolved MR angiography (consensus reading) and DSA were in agreement for nidus size in 23 (82%) of 28 patients (exact CI: 0.631, 0.939). At DSA, seven nidi were smaller than 1 cm, nine were between 1 and 3 cm, and eight were larger than 3 cm. Intermodality agreement between DSA and time-resolved MR angiography was good ({kappa} = 0.76; 95% CI: 0.562, 0.950). The independent raters agreed on the size of the underlying nidus in 24 (86%) of 28 patients (exact CI: 0.673, 0.960), which resulted in very good concordance ({kappa} = 0.80; 95% CI: 0.621, 0.982) (Figs 3, 4).


Figure 3
View larger version (64K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3: Images in 45-year-old man with small AVM (<1 cm) in left temporal lobe. A, Intraarterial DSA image in lateral projection. B, Time-resolved MR angiogram (3.7/1.3, 15° flip angle) reconstructed in sagittal plane from the 3D volume acquired in arterial phase. Images display small nidus (arrow) supplied by a branch of left posterior cerebral artery. (See Movies 5 and 6, http://radiology.rsnajnls.org/cgi/content/full/2463070293/DC1).

 

Figure 4
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4: Images in 16-year-old male patient with medium AVM (1–3 cm) in left parietal lobe. A, Intraarterial DSA image in lateral projection. B, Time-resolved MR angiogram (3.7/1.3, 15° flip angle) reconstructed in sagittal plane from the 3D volume acquired in parenchymal phase. Images display a medium nidus (arrow) supplied by branches of left anterior cerebral artery and middle cerebral artery. AVM is drained via a single dilated cortical vein into the superior sagittal sinus. (See Movies 1 and 2, http://radiology.rsnajnls.org/cgi/content/full/2463070293/DC1).

 
The analysis of venous drainage showed agreement between modalities in 27 (96%) of 28 patients (exact CI: 0.817, 0.999). At DSA, the venous drainage occurred via cortical veins in 18 patients, while in three patients the AVM was drained via the deep venous system, and three AVMs had a mixed venous drainage. The intermodality agreement for venous drainage was excellent ({kappa} = 0.94; 95% CI: 0.814, 1.000). Raters agreed on the venous drainage at time-resolved MR angiography in 25 (89%) of 28 patients (exact CI: 0.718, 0.977) with very good interobserver agreement ({kappa} = 0.83; CI: 0.650, 1.000) (Table 3).


View this table:
[in this window]
[in a new window]

  		Table 3. Interobserver and Intermodality Agreement for AVM Characterization

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 
Our study results suggest that time-resolved MR angiography with parallel imaging in combination with keyhole acquisition and CENTRA k-space sampling techniques is a reliable diagnostic tool for the characterization of intracranial AVMs. The agreement with DSA, which served as the standard of reference, was good for nidus size ({kappa} = 0.76; 95% CI: 0.562, 0.950) and was excellent with respect to arterial supply ({kappa} = 0.91; 95% CI: 0.786, 1.000) and venous drainage ({kappa} = 0.94; 95% CI: 0.814, 1.000).

The common technique of time-of-flight MR angiography has been reported to be unable to adequately depict AVM anatomy (15). There are particular difficulties in demonstrating the correct size of the nidus and draining veins when compared with DSA. The reasons for the shortcomings of time-of-flight MR angiography have been discussed in the literature and include inadequate spatial resolution, spin saturation effects, inadequate signal-to-noise ratio, inadequate background suppression, susceptibility effects, and motion artifacts. Particularly troublesome in AVM imaging with time-of-flight MR angiography is inconsistent depiction of the draining veins and nidus because of overwhelming saturation effects (15).

The clinical use of time-resolved MR angiography has so far been limited by restricted spatial and temporal resolution, as well as a lack of isotropic 3D data. The use of parallel imaging techniques helped improve time-resolved MR angiography by either increasing the temporal resolution at a constant spatial resolution or by allowing high-resolution imaging at an unchanged acquisition time.

Gauvrit et al (12) demonstrated that time-resolved MR angiography with SENSE provides acceptable spatial resolution at 1.5 T. They obtained a spatial resolution of nonisotropic 0.58 x 0.58 x 5 mm (1.68 mm3) after zero filling, with a temporal resolution of 1.7 seconds per volume. However, the imaging slab of 10 cm covered the head only partially, and three different runs with three contrast material injections were necessary to obtain sagittal, frontal, and transverse views.

Summers et al (11) described improved spatial resolution of time-resolved MR angiography with SENSE at 3.0 T. They achieved an in-plane resolution of 0.6 x 1 mm and a reconstruction matrix of 0.45 mm after zero filling at a temporal resolution of 0.5 second per volume. The signal-to-noise ratio of MR angiographic data seemed less optimal because of the extremely high in-plane resolution, which was only in part compensated for by the use of 3.0 T. The solution here might have been the implementation of the keyhole method.

The time-resolved MR angiographic sequence used in our study is based on a combination of SENSE, keyhole acquisition, and CENTRA k-space sampling techniques. The keyhole technique allows accelerated, dynamic MR angiography by restricting repetitive k-space sampling to a predefined central sphere of k-space—representing mainly the information related to the contrast of the MR image—during the first pass of the contrast agent; this enables a temporal resolution of 1.7 seconds per volume. The periphery of k-space—responsible mainly for the spatial resolution—is collected from a reference image (imaging time, 10.9 seconds) obtained at the end of the dynamic acquisition. The information of the center of k-space acquired for each dynamic image is combined with the data from the reference image to obtain a complete data set from every single dynamic volume, which allows the reconstruction of MR images with a high spatial resolution.

The MR imaging sequence includes a keyhole acquisition technique with a central keyhole of 16% of the total k-space, which results in an acceleration factor of 6.25 compared with nonkeyhole imaging. It is the keyhole acquisition in combination with parallel imaging techniques (SENSE factor, eight) that allows highly accelerated contrast-enhanced MR angiography. The additionally applied CENTRA k-space sampling technique is supposed to help avoid artifacts related to fast variations of the contrast agent. With CENTRA, the readout of the midsphere of the kxky-space of the dynamic acquisition is performed in a random order, whereas the readout of the periphery of k-space—sampled from the reference acquisition—is performed in an elliptic manner (16,17). The additional subtraction of the dynamic volumes with a native, noninjected volume enables an exclusive visualization of the cerebral vasculature.

For the practical use of this MR angiographic sequence, it is important to bear in mind that the technique is highly susceptible to patient movement. Because the center and the periphery of k-space are acquired at two different points in time, any patient movement in between results in a disconnection of corresponding k-space information and will cause blurring on the MR images. In addition, digital subtraction with an unenhanced volume may cause suppression of small vessels in case of patient movement.

In contrast to current time-resolved MR angiographic sequences, the combination of SENSE, keyhole acquisition, and CENTRA k-space sampling techniques allows the acquisition of near-isotropic data sets covering the entire head at a millimetric resolution. Its temporal resolution is sufficient to differentiate between early arterial, arterial, parenchymal, and venous phases. On the basis of these near-isotropic 3D data, maximal intensity projections can be calculated, displaying the angioarchitecture of an AVM in any spatial orientation. With the availability of isotropic 3D data in a millimetric resolution, time-resolved MR angiography may soon approach the requisite submillimetric spatial resolution, allowing image data to be integrated into image-guided therapy of intracranial AVMs at radio- or microsurgery (18).

Time-resolved MR angiography is not able to replace DSA in the diagnosis of suspected brain AVMs. The spatial and temporal resolution of the proposed MR angiographic sequence is still insufficient to reliably depict small AVMs, AVMs with a low nidal flow, or intranidal aneurysms and aneurysms of the feeding arteries. We therefore consider the integration of time-resolved MR angiographic data into image-guided radiosurgery of intracranial AVMs and the follow-up of AVMs treated with radiosurgery to be the major indications for this technique.

The spontaneous obliteration of four AVMs in 28 patients with 29 AVMs needs further discussion. Various mechanisms have been implicated for this phenomenon, which has repeatedly been reported in the literature during the years on a case-report basis. Most often, spontaneous regression of AVMs seems to be related to an intracerebral or subarachnoid hemorrhage. The hematoma, with its subsequent mass effects, compresses the draining veins near the nidus rather than the arterial feeders, which causes subsequent thrombosis of the AVM nidus (19). In our series, three of four spontaneous obliterations occurred in patients with intracerebral hemorrhage. Presumably because of the rarity of the disease, with prevalence estimates ranging from five to 631 intracranial AVMs per 100 000 (20), to our knowledge no studies of a larger scale have been published on spontaneous regression of intracranial AVMs as yet.

We acknowledge the following limitations of our study. First, our data were obtained from a single-center study in a small study population; therefore, we cannot make any statement on the equivalence of time-resolved MR angiography when compared with DSA. Second, the time-resolved MR angiographic sequence proposed in this paper has not been compared with other, well-established MR angiographic sequences. Third, the agreement between time-resolved MR angiography and DSA was based on a comparison of supplying vessels, nidus size, and draining veins and did not include the actual number of supplying vessels and draining veins per AVM. Finally, there was a selection bias, in that—because of the setting of our study—only patients who met the selection criteria for gamma knife treatment were included (21). Finally, our study was biased because the raters who reviewed the time-resolved MR angiograms knew they had to expect an AVM.

Because of these inherent limitations, we intend to restrict the clinical application of time-resolved MR angiography to the assessment of nidal response in AVMs treated with radiosurgery in an ongoing study. If the nidus of the AVM does not appear at time-resolved MR angiography performed 12 months after treatment, with the initial time-resolved MR angiograms obtained the day before the radiosurgical treatment serving as comparators, a final DSA examination is performed to confirm complete nidus obliteration. The proposed algorithm may help reduce the radiation dose and the potential risk of an invasive technique in those patients who require repeated angiographic controls (22–27).

In conclusion, agreement (good to excellent) between time-resolved MR angiographic and DSA findings suggests that time-resolved MR angiography is a reliable tool for the characterization of intracranial AVMs with respect to arterial feeders, nidus size, and venous drainage.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
IMPLICATION FOR PATIENT CARE
 References
 


   IMPLICATION FOR PATIENT CARE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
References
 


   ACKNOWLEDGMENTS
 
The authors thank Meinhard Mende, MSc, from the Institute for Medical Biometry, Informatics and Epidemiology of the University of Bonn, Germany, for his help with statistical analysis.


   FOOTNOTES
 

Abbreviations: AVM = arteriovenous malformation • CENTRA = contrast-enhanced robust-timing angiography • CI = confidence interval • DSA = digital subtraction angiography • SENSE = sensitivity encoding • 3D = three-dimensional

Guarantor of integrity of entire study, C.A.T.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, C.A.T., J.G., H.R., N.R., J.Y.G., X.L.; clinical studies, C.A.T., H.R., N.R., X.L., V.L.T.; statistical analysis, C.A.T., J.Y.G.; and manuscript editing, C.A.T., J.G., H.R., J.Y.G., X.L.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
DISCUSSION
ADVANCES IN KNOWLEDGE
 IMPLICATION FOR PATIENT CARE
 References
 

  1. Marchal G, Bosmans H, Van Fraeyenhoven L, et al. Intracranial vascular lesions: optimization and clinical evaluation of three-dimensional time-of-flight MR angiography. Radiology 1990;175:443–448. [Abstract/Free Full Text]
  2. Mehta MP, Petereit D, Turski P, Gehring M, Levin A, Kinsella T. Magnetic resonance angiography: a three-dimensional database for assessing arteriovenous malformations—technical note. J Neurosurg 1993;79:289–293. [Medline]
  3. Bednarz G, Downes B, Werner-Wasik M, Rosenwasser RH. Combining stereotactic angiography and 3D time-of-flight magnetic resonance angiography in treatment planning for arteriovenous malformation radiosurgery. Int J Radiat Oncol Biol Phys 2000;46:1149–1154. [CrossRef][Medline]
  4. Wang Y, Johnston DL, Breen JF, et al. Dynamic MR digital subtraction angiography using contrast enhancement, fast data acquisition, and complex subtraction. Magn Reson Med 1996;36:551–556. [Medline]
  5. Griffiths PD, Hoggard N, Warren DJ, Wilkinson ID, Anderson B, Romanowski CA. Brain arteriovenous malformations: assessment with dynamic MR digital subtraction angiography. AJNR Am J Neuroradiol 2000;21:1892–1899. [Abstract/Free Full Text]
  6. Wetzel SG, Bilecen D, Lyrer P, et al. Cerebral dural arteriovenous fistulas: detection by dynamic MR projection angiography. AJR Am J Roentgenol 2000;174:1293–1295. [Free Full Text]
  7. Aoki S, Yoshikawa T, Hori M, et al. MR digital subtraction angiography for the assessment of cranial arteriovenous malformations and fistulas. AJR Am J Roentgenol 2000;175:451–453. [Free Full Text]
  8. Pruessmann KP, Weiger M, Scheidegger MB, Boesiger P. SENSE: sensitivity encoding for fast MRI. Magn Reson Med 1999;42:952–962. [CrossRef][Medline]
  9. Weiger M, Pruessmann KP, Kassner A, et al. Contrast-enhanced 3D MRA using SENSE. J Magn Reson Imaging 2000;12:671–677. [CrossRef][Medline]
  10. Tsuchiya K, Aoki C, Fujikawa A, Hachiya J. Three-dimensional MR digital subtraction angiography using parallel imaging and keyhole data sampling in cerebrovascular diseases: initial experience. Eur Radiol 2004;14:1494–1497. [Medline]
  11. Summers PE, Kollias SS, Valavanis A. Resolution improvement in thick-slab magnetic resonance digital subtraction angiography using SENSE at 3T. J Magn Reson Imaging 2004;20:662–673. [CrossRef][Medline]
  12. Gauvrit JY, Leclerc X, Oppenheim C, et al. Three-dimensional dynamic MR digital subtraction angiography using sensitivity encoding for the evaluation of intracranial arteriovenous malformations: a preliminary study. AJNR Am J Neuroradiol 2005;26:1525–1531. [Abstract/Free Full Text]
  13. Meckel S, Mekle R, Taschner C, et al. Time-resolved 3D contrast-enhanced MRA with GRAPPA on a 1.5-T system for imaging of craniocervical vascular disease: initial experience. Neuroradiology 2006;48:291–299. [CrossRef][Medline]
  14. van Vaals JJ, Brummer ME, Dixon WT, et al. "Keyhole" method for accelerating imaging of contrast agent uptake. J Magn Reson Imaging 1993;3:671–675. [Medline]
  15. Farb RI, McGregor C, Kim JK, et al. Intracranial arteriovenous malformations: real-time auto-triggered elliptic centric-ordered 3D gadolinium-enhanced MR angiography—initial assessment. Radiology 2001;220:244–251. [Abstract/Free Full Text]
  16. Willinek WA, Gieseke J, Conrad R, et al. Randomly segmented central k-space ordering in high-spatial-resolution contrast-enhanced MR angiography of the supraaortic arteries: initial experience. Radiology 2002;225:583–588. [Abstract/Free Full Text]
  17. Hoogeveen RM, von Falkenhausen M, Gieseke J. Fast, dynamic high resolution contrast-enhanced MR angiography with CENTRA keyhole and SENSE [abstr]. In: Proceedings of the 12th Meeting of the International Society for Magnetic Resonance in Medicine. Berkeley, Calif: International Society for Magnetic Resonance in Medicine, 2004.
  18. Taschner CA, Le Thuc V, Reyns N, et al. Gamma knife surgery for arteriovenous malformations in the brain: integration of time-resolved contrast-enhanced magnetic resonance angiography into dosimetry planning. J Neurosurg 2007;107(4):854–859. [Medline]
  19. Krapf H, Siekmann R, Freudenstein D, Küker W, Skalej M. Spontaneous occlusion of a cerebral arteriovenous malformation: angiography and MR imaging follow-up and review of the literature. AJNR Am J Neuroradiol 2001;22:1556–1560. [Abstract/Free Full Text]
  20. Berman MF, Sciacca RR, Pile-Spellman J, et al. The epidemiology of brain arteriovenous malformations. Neurosurgery 2000;47:389–396. [CrossRef][Medline]
  21. Fleetwood IG, Steinberg GK. Arteriovenous malformations. Lancet 2002;359:863–873. [CrossRef][Medline]
  22. Heiserman JE, Dean BL, Hodak JA, et al. Neurologic complications of cerebral angiography. AJNR Am J Neuroradiol 1994;15:1401–1407. [Abstract]
  23. Cloft HJ, Joseph GJ, Dion JE. Risk of cerebral angiography in patients with subarachnoid hemorrhage, cerebral aneurysm, and arteriovenous malformation: a meta-analysis. Stroke 1999;30:317–320. [Abstract/Free Full Text]
  24. Willinsky RA, Taylor SM, TerBrugge K, Farb RI, Tomlinson G, Montanera W. Neurologic complications of cerebral angiography: prospective analysis of 2899 procedures and review of the literature. Radiology 2003;227:522–528. [Abstract/Free Full Text]
  25. Derdeyn CP, Moran CJ, Eichling JO, Cross DT 3rd. Radiation dose to patients and personnel during intraoperative digital subtraction angiography. AJNR Am J Neuroradiol 1999;20:300–305. [Abstract/Free Full Text]
  26. Ron E. Cancer risks from medical radiation. Health Phys 2003;85:47–59. [CrossRef][Medline]
  27. Berrington de González A, Darby S. Risk of cancer from diagnostic X-rays: estimates for the UK and 14 other countries. Lancet 2004;363:345–351. [CrossRef][Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow AVI Movies
Right arrow All Versions of this Article:
2463070293v1
246/3/871    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Taschner, C. A.
Right arrow Articles by Leclerc, X.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Taschner, C. A.
Right arrow Articles by Leclerc, X.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE